Cutaneous and Ocular Toxicology, 26: 45–56, 2007
ISSN: 1556-9527 print=1556-9535 online
ABSORPTION OF LAWSONE THROUGH HUMAN SKIN1
Margaret E. K. Kraeling and Robert L. Bronaugh
Office of Cosmetics and Colors, US Food and Drug Administration,
College Park, MD
Connie T. Jung
Office of the Commissioner, Office of Policy, US Food
and Drug Administration, Rockville, MD
Lawsone (2-hydroxy-1,4-naphthoquinone) is the principal color ingredient in henna, a color
additive approved with limitations for coloring hair by the Food and Drug Administration
(FDA) under 21 CFR 73.2190. In 2002, the scientific committee on cosmetics and non-food
products (SCCNFP), now known as the scientific committee for consumer products
(SCCP), evaluated the safety of lawsone as a coloring agent in hair dye products of the
European Union (EU). The SCCNFP concluded that lawsone was mutagenic and not suitable
for use as a hair coloring agent. As a result, studies were conducted to measure the
extent of lawsone absorption through human skin. Lawsone skin absorption was determined
from two hair coloring products and two shampoo products, all containing henna. [14C]-
Lawsone (sp. act. 22.9 mCi/mmol) was added to each commercial product and the products
were applied to dermatomed, nonviable human skin mounted in flow-through diffusion cells
perfused with a physiological buffer (HEPES-buffered Hanks’ balanced salt solution, pH
7.4). Products remained on the skin for 5 minutes (shampoos) and 1 hour (hair color
paste). For the henna hair paste products, 0.3 and 1.3% of the applied dose was absorbed
into the receptor fluid in 24 hours while 2.2 and 4.0% remained in the skin. For both henna
shampoo products, 0.3% of the applied dose was absorbed into the receptor fluid at 24 hours
while 3.6 and 6.8% remained in the skin. For all products, most of the lawsone applied was
washed from the surface of the skin (83–102%) at the end of the exposure period. Extended
absorption studies were conducted for 72 hours to determine if skin levels of lawsone in the
24 hour studies might eventually be percutaneously absorbed. These studies determined that
the majority of the lawsone remained in the skin with only a small but significant increase
(for three out of four products) in receptor fluid values. Therefore, it appears that receptor
fluid values would give a good estimate of lawsone absorption for an exposure estimate and
that skin levels of lawsone need not be included.
Keywords: Human skin; Lawsone; Percutaneous absorption
1 The opinions and conclusions expressed in this article are solely the views of the authors and do
not necessarily reflect those of the Food and Drug Administration. This study was conducted under the
Guidelines of Good Laboratory Practice.
This article not subject to United States copyright law.
Address correspondence to Margaret E. K. Kraeling, Office of Cosmetics and Colors, US Food
and Drug Administration, Harvey W. Wiley Building, Room 4E021, 5100 Paint Branch Parkway, College
Park, MD 20740. E-mail: firstname.lastname@example.org
Lawsone (2-hydroxy-1,4-naphthoquinone) is the principal natural dye contained
in the leaves of henna (Lawsonia inermis) at a concentration of about 1.0–1.4% (1).
Because of its natural origin, henna as well as its dye ingredient lawsone have been used
for centuries as a hair dye, body paint, and tattoo dye (1,2). Consumer products currently
containing henna include shampoos, conditioners, hair coloring powders and
creams, body lotions, and body cleansers. However, henna has only been approved with
limitations for coloring hair by the Food and Drug Administration under 21 CFR
73.2190 (3). Henna is not approved for use as a body paint and tattoo dye.
The safety of lawsone as a coloring agent in hair dye products was evaluated by
the scientific committee on cosmetics and non-food products (SCCNFP), now
known as the scientific committee for consumer products (SCCP) of the European
Union. In 2002, the SCCP found that lawsone was mutagenic and clastogenic both
in vitro and in vivo in some experiments (4). Lawsone induced mutations and
chromosomal aberrations in mammalian cells in vitro and had some positive results
in bone marrow micronucleus assays. Based on the available studies, the SCCP concluded
that lawsone was not suitable for use as a hair coloring agent (4). In a 2005
review of lawsone, the SCCP concluded that the available mutagenicity data did not
rule out a clastogenic potential and that the data was currently insufficient to assess
the safe use of lawsone as a hair dye (5).
Because of its wide use as a natural dye ingredient in hair products, we initiated
studies to measure the extent of lawsone absorption through human skin. [14C]-
Lawsone was added to four commercially available henna-containing pastes and
shampoo products and applied to human skin in diffusion cells under conditions that
simulated cosmetic consumer use.
MATERIALS AND METHODS
14C]-Lawsone (2-hydroxy-[1-14C]-4-naphthoquinone; specific activity 22.9 mCi=
mmol; 93–97% radiochemical purity) was synthesized by Research Triangle Institute
(Research Triangle, NC). [3
H]-Water (specific activity 55.5 mCi=mmol; 97% purity)
was obtained from New England Nuclear Corp. (Boston, MA). Nonradiolabeled lawsone
and a lawsone metabolite (2, 2-Bi(3-hydroxy)-1,4-naphthoquinone) were
purchased from Sigma Chemical Co. (St. Louis, MO). High performance liquid chromatography
(HPLC) grade solvents were obtained from J.T. Baker Chemical Co.
(Phillipsburg, NJ). HEPES-buffered Hanks’ balanced salt solution (HHBSS; Hanks’
balanced salt solution dry powder packets prepared by Gibco BRL, Life Technologies,
Grand Island, NY) was prepared fresh prior to each study. The commercially available
henna hair powders=pastes (formulations A and B) and henna shampoos (formulations C and D)
containing lawsone were obtained from a local cosmetics supplier. Analysis and Purification Method for Lawsone
Purity of radiolabeled lawsone was determined by thin layer chromatography (TLC) methodology. An aliquot of [14C]-lawsone was spotted on a silica TLC plate 46 M. E. K. KRAELING ET AL. and the plate developed in a tank containing 80% chloroform and 20% methanol as the solvent system. The location and the amount of radioactivity was determined with a Bioscan System 200 and System A-2000 Imaging TLC plate scanner (Bioscan, Washington, DC).
Because of a continual slow decrease in the purity of [14C]-lawsone, aliquots from the radiochemical stock solution were repurified to at least 93% purity prior to each experiment. [14C]-Lawsone was chromatographed on a TLC plate with an unlabeled lawsone and metabolite (2, 2-bi(3-hydroxy)-1,4-naphthoquinone) standard.
The silica at the spot corresponding to lawsone was scraped off the plate, collected, and extracted twice with chloroform:methanol (95:5). The supernatant was pooled and dried under nitrogen gas. The residue was reconstituted in a volume of ethylene glycol dimethyl ether and stored under nitrogen at –80 2
C until used. The purity of the lawsone was analyzed by the TLC method previously described.
Percutaneous Absorption Experiments and Analyses In vitro percutaneous absorption studies were conducted based on methods described in detail for human skin in flow-through diffusion cells (6–8). Lawsone skin absorption studies were conducted using human abdominal skin obtained from the National Disease Research Interchange (NDRI) (Philadelphia, PA). Full-thickness frozen skin from NDRI was stored at 80 2C until use. The skin was thawed to room temperature, subcutaneous fat was removed and the skin surface was gently washed with a 10% (v=v) liquid detergent (Palmolive1 ‘‘Original’’ formula) solution. The skin was cut with a Padgett dermatome (Padgett Instruments, Dermatome Division, Kansas City, MO) to a thickness of approximately 200–300 mm (9). Skin discs were prepared with a punch and placed epidermal side up in teflon flow-through diffusion cells (6). The skin in the diffusion cells was perfused with HHBSS, pH 7.4, receptor fluid at a flow rate of approximately 1.5 mL=hr for the duration of the study (8) with the skin surface temperature of the stratum corneum maintained at 32C. The surface area of skin in the diffusion cell was 0.64 cm2 and the receptor well volume was 260 mL.
Prior to application of the test article, the integrity of the barrier of skin in each diffusion cell was verified by a 20-minute exposure to [3H]-water and subsequent
determination of the percent absorption of the radiolabel (10). Cells in which the percent
of the applied dose of [3
H]-water absorbed through skin was greater than the historical limit of 0.35% were discarded (10).
After application of the test article, diffusion cell receptor fluid fractions were collected at 6-hour intervals for 24 hours using a fraction collector. An aliquot of receptor fluid effluent was removed and the amount of radioactivity in each fraction was determined by liquid scintillation counting (LSC; Beckman LS 6500, Fullerton,
CA). For products A and B (powders=pastes), the skin surface was rinsed 3–5 times with 0.1 mL distilled water 1 hour after application. For products C and D (shampoos), the skin surface was rinsed three times with 0.1 mL distilled water 5 minutes after application. During the rinsing procedure, the skin surface was gently rubbed with cotton-tipped applicators to remove all the distilled water. The cotton tips were collected and the radioactivity was measured by LSC. The skin discs were removed
ABSORPTION OF LAWSONE THROUGH HUMAN SKIN 47
from the diffusion cell and the stratum corneum was tape-stripped 10 times to determine amount of [14C]-lawsone located in the stratum corneum. The remaining
epidermal=dermal tissue in each skin disc was dissolved in Scintigest1 tissue solubilizer.
The amount of radioactivity in the stratum corneum and remaining skin was measured by LSC.
The caps of the diffusion cells were soaked in a 1% soap solution to remove any [14C]-lawsone formulation remaining on the diffusion cell. Aliquots were measured by LSC and included with the rinse values.
In order to determine the fate of the lawsone that remained in the skin at the end of a 24-hour study, in vitro percutaneous absorption studies were extended to 72 hours. The skin in the diffusion cells was rinsed at the specified exposure time according to procedures described above. Receptor fluid samples were collected every 6 hours up to 72 hours. In the extended studies, a [3H]-water skin barrier integrity test of the skin in the diffusion cells was performed at the end of the 72-hour studies to verify barrier integrity. After the barrier integrity test, the stratum corneum was separated by tape stripping and the remaining epidermis=dermis was digested and analyzed as previously described.
Preparation and Application of Test Articles
Two commercially available hair coloring powders containing henna and lawsone were prepared with [14C]-lawsone (about 0.12–0.24 mCi=dose) and applied to skin in an amount that simulated consumer exposure.
Product A was a 60 g package of powder that was mixed with water to form a thick paste to be applied to the hair. Initially, the amount of powder to apply to skin in each diffusion cell was determined. It was assumed that half a package (30 g) was used
for average length hair. The scalp was assumed to have a surface area of 650 cm2 (11), and it was estimated that 10% of the applied product reached the scalp (12). The surface area of exposed skin in the diffusion cell was 0.64 cm2. Therefore the amount of powder applied per cell was: 30=650 0.1 0.64 1000 mg=g ¼ 3.0 mg. Enough paste was prepared for twenty doses in each experiment by mixing 60 mg of product A powder with 200 mL of water (containing [14C]-lawsone). A thick paste was formed weighing
260 mg. Therefore, 13.0 mg of paste was applied to skin in each diffusion cell. The
dose of [14C]-lawsone applied in the henna paste was about 0.21 0.03 mCi
(1.51 0.22 mg=cm2) in the 24-hour studies and 0.19 0.004 mCi (1.36 0.03 mg=cm2)
in the 72-hour studies as determined from aliquots taken from the paste.
Product B was an 80 g package of powder that was mixed with water to form a
thick paste to be applied to the hair. Using the same assumptions made for Product
A, the amount of powder needed per cell was 3.9 mg. Using the procedures given for
product A, it was determined that 13.9 mg of paste made with 200 mL of water would
be applied to skin in each diffusion cell. The dose of [14C]-lawsone applied was
approximately 0.23 0.001 mCi (1.65 0.01 mg=cm2) in the 24-hour studies and
0.13 0.01 mCi (0.93 0.04 mg=cm2) in the 72-hour studies.
In case of lawsone containing shampoos, two commercially available henna
shampoo products were prepared with [14C]-lawsone and applied to skin in an
amount that simulated consumer exposure. For each shampoo application, we
assumed a cosmetic, toiletry and fragrance association (CTFA) survey amount of
12.02 g shampoo used=day (13), 650 cm2 surface area of the scalp, and 10% of the
product reaching the scalp. The amount of each shampoo applied to skin in the
diffusion cell was 12.02=650 0.1 0.64 1000 mg=g ¼ 1.2 mg. For 20 doses per
experiment, 24 mg of shampoo was diluted 1 : 6 with water (120 mL) to simulate
shampoo application to wet hair. The final solution was prepared with [14C]-lawsone
and 7.2 mg was applied to skin in each diffusion cell. The actual dose of [14C]-
lawsone applied to each cell was about 0.15 0.01 mCi (1.08 0.07 mg=cm2) (product
C) and 0.14 0.04 mCi (1.00 0.29 mg=cm2) (product D) in the 24-hour studies and
0.11 0.02 mCi (0.79 0.14 mg=cm2) (product C) and 0.09 0.003 mCi
(0.65 0.02 mg=cm2) (product D) in the 72-hour studies.
Data and Statistical Analysis
The results from the lawsone penetration analyses are given in Tables 1–4. The
sum of the total amount of lawsone measured in skin layers and receptor fluid at the
end of the 24 and 72-hour experiments was defined as total lawsone penetrated.
Results are given for the mean SEM from replicate measurements of one to three
donors and expressed as the percent of the applied dose of total lawsone penetrated.
Absorption is defined as the amount of test compound appearing in the receptor
fluid. Statistical analyses of differences between the 24 and 72-hour data were conducted
using Student’s t-test (p < 0.05) (SigmaStatTM Statistcal Software, Jandel Scientific Software, San Rafael, CA).
Commercially available henna hair powder=paste formulations containing
[14C]-lawsone were applied to excised human skin in diffusion cells. The total percent
of applied lawsone penetrating nonviable human skin from consumer product A in
24 hours was 2.5% (Table 1). Only 0.3% was absorbed into the receptor fluid. For
consumer product B, total percent lawsone penetration after 24 hours was higher at
5.3% (Table 2). Considerably more lawsone was absorbed into the receptor
Substantial amounts of lawsone (75–88% of penetrated dose) remained in the
skin at the end of the 24-hour studies with henna powder=paste formulations.
Approximately two-thirds of the lawsone remaining in the skin was found in
the stratum corneum (Fig. 1). Therefore, extended studies were conducted that continued
for an additional 48 hours (72 hours total) to determine if any additional
lawsone would diffuse into the receptor fluid from skin. For consumer product A,
total lawsone remaining in the skin decreased significantly to 0.8% (p ¼ 0.004),
while the amount of lawsone absorbed into the receptor fluid (0.5%) increased by
a small but significant amount (Table 1). For consumer product B, total lawsone
remaining in the skin decreased to 2.9%, while the amount of lawsone absorbed into
the receptor fluid (1.6%) did not increase significantly (Table 2).
The 72-hour absorption rate profile of lawsone through human skin following
the 1-hour application of the henna hair paste formulations was obtained by plotting
the percent of applied dose absorbed in each 6-hour receptor fluid fraction versus.
time (Fig. 2). The highest rate of lawsone absorption occurred during the first
6 hours, followed by a decrease in the absorption rate to near baseline levels at
18 hours for product A and 30 hours for product B.
Lawsone human skin penetration was also determined from commercially
available henna shampoo formulations applied on skin for 5 minutes. The total percent
of applied lawsone penetrated from consumer product C in human skin in 24
hours was 3.9% (Table 3). Only 0.3% was absorbed into receptor fluid. For consumer
product D, total percent lawsone penetration after 24 hours was the highest
of all products at 7.2% (Table 4). A small amount of lawsone (0.3%) was absorbed
into the receptor fluid. Substantial amounts of lawsone from products C and D (92
to 94% of the penetrated dose) remained in the skin at the end of the 24-hour
studies. As with the hair coloring products, approximately two-thirds of the amount
found in skin was located in the stratum corneum (Fig. 3).
Extended (72 hours) studies were also conducted for the henna shampoo
formulations because of the substantial amount of lawsone remaining in skin at
24 hours. For product C, total lawsone remaining in the skin at 72 hours decreased
significantly to 1.9% (p ¼ 0.007) (lawsone levels in the stratum corneum and epidermis=dermis
both decreased significantly), while the receptor fluid value at 72 hours
increased by a small but significant amount from the 24-hour value.
Extended studies for consumer product D found a decrease in skin levels of lawsone
to 3% (not significant) and a significant increase in lawsone absorption into the
receptor fluid to 0.6% (p ¼ 0.015) when the 24 and 72 hours values were compared.
The 72-hour absorption rate profiles of lawsone through human skin from the
shampoo formulations (products C and D) are shown in Fig. 4. Again, the highest
rate of lawsone absorption occurred during the first 6 hours
In all of the extended studies, the integrity of the barrier of the skin in the diffusion
cells after 72 hours was measured by performing a [3
H]-water skin barrier
integrity test. This was done to verify that any decrease in lawsone values in skin
were not because of degradation of the skin’s barrier. Penetration of [3
24 hours was about 0.13%, while penetration at 72 hours was about 0.11%. It
appeared that the barrier of the skin samples in the diffusion cells did not break
down during the 72 hours extended studies.
Although henna is used in a number of cosmetic products, there is limited
information on the exposure of consumers to henna in these products. Lawsone,
the active ingredient of henna, has been found in this study to be absorbed through
human skin when applied in commercially available henna containing cosmetic products.
Absorption into the diffusion cell receptor fluid during 24-hour studies ranged
from 0.3–1.3% of the applied dose after short-term skin exposure to the four hair
coloring and shampoo products. On an average, about 90% of the lawsone that
penetrated into the skin from these products during the absorption studies remained
there at the end of 24 hours. Approximately two-thirds of this material in the skin
resided in the stratum corneum.
Recently, the percutaneous absorption of lawsone has been evaluated in
human skin by addition of lawsone to a commercial hair dye formulation (1). The
formulation (containing 2.04% lawsone) was applied at 20 mg=cm2 for 30 minutes
on dermatomed human skin in static diffusion cells. After 24 hours, 0.49 0.39%
of the applied lawsone was absorbed into receptor fluid with 0.14 0.04% remaining
in the stratum corneum and 0.13 0.06% in the epidermis=dermis. The majority
of the lawsone appeared to penetrate through the skin into the receptor fluid; however,
the receptor fluid was not identified (1). The percent of applied lawsone
absorbed (0.49%) found by Kirkland and Marzin was similar to values reported
in our study and was obtained by applying lawsone in a 20 mg dose of hair dye formulation
per cm2 skin. We had chosen similar amounts of product to apply to
skin using a 20 mg dose of henna paste and an 11.3 mg dose of henna shampoo
per cm2 skin.
A key issue in determining possible systemic absorption of lawsone is the fate
of the compound remaining in the skin at the end of the study. We have previously
used an extended absorption study to estimate the amount of material in the skin at
the end of a study that may be eventually systemically absorbed (14–16). With the
lipophilic fragrance musk xylol (14), skin absorption into the receptor fluid doubled
from approximately 6.5% to 13% when studies were extended to 72 hours. This
result suggested that substantial amounts of musk xylol in skin at the end of the
24-hour study would eventually be systemically absorbed. A skin reservoir was also
shown to be formed with the more polar chemicals glycolic and lactic acid (7). When
applied in a pH 3.0 formulation, 2.6% of the applied dose of glycolic acid was
absorbed through the skin while 24.6% remained in the skin at the end of the
24-hour studies. A large skin reservoir was also formed in 24-hour absorption studies
with the diethanolamine (DEA) (15), and dihydroxyacetone (16). For these compounds
it appeared that most of the penetrated material in the skin at 24 hours
54 M. E. K. KRAELING ET AL.
remained in this skin reservoir during the extended studies and was unavailable for
The recently issued test guideline by Organisation for Economic Co-operation
and Development (OECD) for percutaneous absorption studies (17) state that skin
levels must be included as being systemically absorbed unless additional studies are
conducted to demonstrate a lack of absorption. In the current study, we have conducted
extended studies with the lawsone commercial products to help resolve this
issue. In the additional 48 hours following the 24-hour study, significant (three
out of four compounds) but small increases in absorption of lawsone were observed.
The small increases in absorption observed at 72 hours suggest that the receptor fluid
levels of lawsone are a reasonable estimate of systemic absorption and that skin
levels need not be included as potential systemically available material.
Another approach for evaluating possible systemic absorption of compounds in
the skin reservoir was recommended by the SCCP (18). Their approach was to include
only the compound that had penetrated into the viable epidermis and dermis as potentially
systemically absorbed and to add these values to receptor fluid levels to obtain a
value for systemic absorption. Compound that was found in the stratum corneum layer
was not included as systemically absorbed. This approach also has merit as a way to
approximate the systemic fate of a compound remaining in the skin at the end of a study.
Kirkland and Marzin included the epidermal=dermal levels of lawsone as being systemically
absorbed but only a small amount of lawsone was found in that region.
In the current study, most of the material remaining in the skin was in the stratum
corneum. So with the SCCP approach to estimate systemic absorption, most of
the skin reservoir of lawsone would not be included as absorbed. A similar estimate
of systemic absorption is, therefore, obtained from the results of the extended study
that suggests the use of only receptor fluid values for this estimate.
Lawsone has been found to be absorbed through human skin into the receptor
fluid of diffusion cells from both henna hair paste and shampoo products. The
majority of the applied lawsone that had penetrated the skin remained in the skin
with most of the lawsone localized in the stratum corneum. In in vitro percutaneous
absorption studies where a substantial amount of test chemical remains in the skin at
the end of a 24-hour study, it is important to evaluate the systemic fate of this
material in the skin before an estimation of systemic absorption is determined. If
the systemic fate of material in the skin is not adequately estimated an inaccurate
determination of skin absorption will be made.
1. Kirkland D, Marzin D. An assessment of the genotoxicity of 2-hydroxy-1,4-naphthoquinone,
the natural dye ingredient in henna. Mut Res 2003; 537:183–199.
2. Nohynek GJ, Fautz R, Benech-Kieffer F, Toutain H. Toxicity and human health risk of
hair dyes. Food Chem Tox 2004; 42:517–543.
3. U.S. Code of Federal Regulations, 1982, Title 21, Section 73.2190.
4. The Scientific Committee for Cosmetics and Non-Food Products (SCCNFP): Opinion of
the SCCNFP concerning Lawsone. 2002.
5. The Scientific Committee on Consumer Products (SCCP): Opinion of the SCCP on
Lawsonia inermis (Henna). 2005.
ABSORPTION OF LAWSONE THROUGH HUMAN SKIN 55
6. Bronaugh RL, Stewart RF. Methods for in vitro percutaneous absorption studies IV: The
flow-through diffusion cell. J Pharm Sci 1985; 74:64–67.
7. Kraeling MEK, Bronaugh RL. In vitro percutaneous absorption of alpha hydroxy acids
in human skin. J Soc Cos Chem 1997; 48:187–197.
8. Collier SW, Sheikh NM, Sakr A, Lichtin JL, Stewart RF, Bronaugh RL. Maintenance of
skin viability during in vitro percutaneous absorption=metabolism studies. Toxicol
Appl Pharmacol 1989; 99:522–533.
9. Bronaugh RL, Stewart RF. Methods for in vitro percutaneous absorption studies VI:
Preparation of the barrier layer. J Pharm Sci 1986; 75:487–491.
10. Bronaugh RL, Stewart RF, Simon M. Methods for in vitro percutaneous absorption
studies VII: Use of excised human skin. J Pharm Sci 1986; 75:1094–1097.
11. Yourick JJ, Bronaugh RL. Percutaneous penetration and metabolism of 2-nitro-p-phenylenediamine
in human and fuzzy rat skin. Tox App Pharm 2000; 166:13–23.
12. European Centre for Ecotoxicology and Toxicology of Chemicals (ECETOC) Technical
Report No. 58. Assessment of Nonoccupational Exposure to Chemicals, 1994:20–21.
13. Cosmetic, Toiletry and Fragrance Association (CTFA). 1983. Summary of the results of
surveys of the amount and frequency of use of cosmetic products by women. Prepared by
Environ Corporation, Washington, DC for CTFA., Washington, DC. 1–7.
14. Hood HL, Wickett RR, Bronaugh RL. The in vitro percutaneous absorption of the
fragrance ingredient musk xylol. Fd Chem Tox 1996; 34:483–488.
15. Kraeling MEK, Yourick JJ, Bronaugh RL. In vitro human skin penetration of diethanolamine.
Fd Chem Tox 2004; 42:1553–1561.
16. Yourick JJ, Koenig ML, Yourick DL, Bronaugh RL. Fate of chemicals in skin after dermal
application: does the in vitro skin reservoir affect the estimate of systemic absorption?
Tox Appl Pharm 2004; 195:309–320.
17. Anonymous, OECD Guideline for the Testing of Chemicals: Skin Absorption (in vitro
method), Organization for Economic Cooperation and Development, Paris, 2004; 1–8.
18. Anonymous. Basic criteria for the in vitro assessment of dermal absorption of cosmetic
ingredients, updated 2003, Scientific Committee on Cosmetic Products and Non-Food
Products Intended for Consumers, European Commission, 2003; 1–9.